Share New Research 
@rakinmunim sent me a paper a while ago that I finally had the time to figure out. I'm going to paste the readme on my github here, because it is pretty short and sweet and to the point.
https://github.com/algo-dude/QuantConnect_public/tree/VWAP_trend
There is a popular SSRN paper floating around the quant world from some big names in the industry. You can find the paper here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4631351
I am fairly skeptical about claims such as
- Our findings reveal that an initial investment of $25,000 in the VWAP Trend Trading strategy with QQQ would have grown to $192,656, net of commissions, yielding a 671% return. This performance is marked by a maximum drawdown of just 9.4% and a Sharpe Ratio of 2.1.
My personal opinion is that profitable research on trading strategies is extremely unlikely to be made public. I have attempted to recreate this paper's methodology with QuantConnect.
You can see a full backtest here: https://www.quantconnect.com/reports/2dd44a5c1394e0ba3de9c7f4d7dc0716
You can clone the backtest here: https://www.quantconnect.com/terminal/clone/-/2dd44a5c1394e0ba3de9c7f4d7dc0716/clone-of%3A-emotional-fluorescent-yellow-seahorse
Let me be very clear here unlike on my github: I wrote and tested this to publicly refute these guys that run a trading firm. I saw the owner of Bear Bull Traders published something else recently, and I think since I am new in the industry, it wouldn't hurt to write some counters to their paper and their claims. Check the paper and check the backtest code. If there's a glaring omission on my part, please let me know before I put them on blast on my LinkedIn.
**TLDR: I do not know why this is a published paper. If my logic is wrong, please comment here, ping me on Discord, or raise a PR.**
Mia Alissi
Based on the information provided in the extracts, it seems 'Algo_dude' is seeking feedback on his implementation of a trading strategy from a paper he read. He has shared his code on GitHub and run a backtest on QuantConnect, and he's asking for any potential errors or omissions to be pointed out before he publicly refutes the claims made in the paper.
Unfortunately, the extracts provided do not contain any specific details about 'Algo_dude's' implementation or the paper's methodology, so it's not possible to provide specific feedback or suggestions for improvement.
It's recommended that 'Algo_dude' shares more specific details about the trading strategy and his implementation, such as the trading signals, the assets being traded, the trading frequency, and any risk management techniques used. This would make it easier for others to provide constructive feedback.
In general, it's a good practice to thoroughly test any trading strategy with both backtesting and paper trading before using it for live trading. This includes testing the strategy with different market conditions and assumptions to ensure its robustness.
If 'Algo_dude' has any specific questions about using QuantConnect or the LEAN engine, he might find the QuantConnect Documentation helpful. It provides detailed information about the platform's features and how to use them, including data import, universe selection, alpha creation, portfolio construction, execution, and risk management.
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
Algo_dude
Their paper:
My implementation:
Jared Broad
This is a simplified view of your implementation – but it looks like the minimum isn't doing what it intended. What was the goal of the minimum? (Didn't read paper 😬)
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
Jared Broad
It's pretty sensitive to the minimum filter and gets eaten alive by covid. Tweaks to make the variables more apparent.
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
Algo_dude
Thanks Jared. I actually added the minimum after emailing the guys who wrote the paper and trying to make my code more similar to the MATLAB code that he referred to. I think this is what I was going for:
So, we don't trade if the difference between price and VWAP is too small - just wait until it is larger.
Sebastien M. Laignel
Hi there,
I have also significanty improved the very poor performance of this algo by adding a 2% deadband across the VWAP, but I am far from approaching the original paper results. My opinion is that there is a major failure on the MATLAB code. Has that enigma been fully elucidated already ?
Brgds
Algo_dude
The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.
To unlock posting to the community forums please complete at least 30% of Boot Camp.
You can continue your Boot Camp training progress from the terminal. We hope to see you in the community soon!